[mkgmap-dev] Dead End Check

Thu Aug 30 18:06:37 BST 2012

Nope, you and others need to be more carefull!!!

access=private doesn't only say that it's a private street, but that you 
may not use it now matter how. If you're allowed to use it by foot - 
then add foot=yes/permissive, respectively for bicycle or other stuff. 
Of course we would need to add those checks. But for a plain 
highway=service/footway/track/path/road with access=private - that 
usually shouldn't be in a map at all, and definitely not routable 
(knowing access restrictions don't work on Basecamp 3.3 or Garmin GPS 
with activity routing).

We have not key at all in OSM to say that a road is private!
On 30.08.2012 17:31, Geoff Sherlock wrote:
> On 30.08.2012 16:59, Felix Hartmann wrote regarding micromapping:
>
> "Plus access=no
> / access=private and length is up to 50m then delete it"
>
> Felix, I have certainly taken private roads > 50m that led to a pubic 
> footpath (what you may describe as unpaved track); and the private 
> road was open to pedestrians to use. So I think more care may be 
> needed here - although I don't know what the laws are in 
> Germany/Austria regarding private roads.
>
> Geoff.
>
> -----Original Message----- From: Felix Hartmann
> Sent: Thursday, August 30, 2012 3:59 PM
> To: Development list for mkgmap
> Subject: Re: [mkgmap-dev] Dead End Check
>
>
> On 30.08.2012 16:48, Marko Mäkelä wrote:
>> I think that a more severe problem are routing islands, because they
>> really break routing. Some mappers seem to think that a
>> highway=pedestrian area is enough and may even delete an overlapping
>> highway=* way as garbage.
>>
>> As far as I know, currently mkgmap does not check for routing islands.
>> It might be hard to avoid false positives near tile (or map extract)
>> borders. If we ignore that, is it computationally expensive to detect
>> routing islands? (Run Tarjan's algorithm and complain if there are
>> several strongly connected components in the routing graph, perhaps?)
> Oh I also don't think they mess up with routing (at least not much). But
> they do visually mess up the map, and take data/space that is not
> needed. If this micromapping takes overhand, then at some point, 50% of
> all footways are rubbish in Austria. Outside of Germany and Austria such
> problems virtually not exist, but it seems here more and more mappers
> turn crazy and make Openstreetmap a cadastral map. A simple check for if
> lenght is under 10m and it's a dead end, then delete it. Plus access=no
> / access=private and length is up to 50m then delete it, would be enough
> to solve 90% of that cases, without any negative impacts (maybe a couple
> of wrongly tagged streets disappear, however if the 50m and access=no //
> access=private behaviour were in the mkgmap default style, I could well
> believe those tagging errors to be corrected quickly, and wrong new ones
> corrected quickly too).
>
> Routing Islands are another point. I would love to have mkgmap search
> for the nearest routable way at each end of a ferry and
> gondolas/chairlifts which is dead ended and connect it via straight line
> (using 0x08 in my case, as I set that type to be invisible and as a
> "highway link" it would give best routing descriptions).
>> Best regards,
>>
>> Marko
>> _______________________________________________
>>
>

-- 
keep on biking and discovering new trails

Felix
openmtbmap.org & www.velomap.org